Fast and automatic estimation of transition state structures using tight binding quantum chemical calculations
Author(s) -
Maria H. Rasmussen,
Jan H. Jensen
Publication year - 2020
Publication title -
peerj physical chemistry
Language(s) - English
Resource type - Journals
ISSN - 2689-7733
DOI - 10.7717/peerj-pchem.15
Subject(s) - transition state theory , quantum , path (computing) , false positive paradox , tight binding , transition state , binding energy , chemistry , statistical physics , physics , computational chemistry , electronic structure , atomic physics , quantum mechanics , mathematics , computer science , statistics , reaction rate constant , kinetics , biochemistry , programming language , catalysis
We present a method for the automatic determination of transition states (TSs) that is based on Grimme’s RMSD-PP semiempirical tight binding reaction path method (J. Chem. Theory Comput. 2019, 15, 2847–2862), where the maximum energy structure along the path serves as an initial guess for DFT TS searches. The method is tested on 100 elementary reactions and located a total of 89 TSs correctly. Of the 11 remaining reactions, nine are shown not to be elementary reactions after all and for one of the two true failures the problem is shown to be the semiempirical tight binding model itself. Furthermore, we show that the GFN2-xTB RMSD-PP barrier is a good approximation for the corresponding DFT barrier for reactions with DFT barrier heights up to about 30 kcal/mol. Thus, GFN2-xTB RMSD-PP barrier heights, which can be estimated at the cost of a single energy minimisation, can be used to quickly identify reactions with low barriers, although it will also produce some false positives.
Accelerating Research
Robert Robinson Avenue,
Oxford Science Park, Oxford
OX4 4GP, United Kingdom
Address
John Eccles HouseRobert Robinson Avenue,
Oxford Science Park, Oxford
OX4 4GP, United Kingdom